Dissection of anti-CTLA4-induced cytotoxic T-cell responses in melanoma.

9078 Background: There is strong evidence that melanoma-reactive T cells induced by immunotherapeutic interventions such as anti-CTLA4 therapy can exert clinically effects. However, there is very little information on how these therapies influence tumor-specific T cell responses. Furthermore, as the number of potential melanoma-associated antigens to which these responses can be directed is very high, classical strategies to map cytotoxic T cell reactivity do not suffice. Knowledge of such reactivities would be useful to design targeted strategies, selectively aiming to induce immune reactivity against these antigens. Methods: We have addressed these issues by designing MHC class I molecules occupied with UV-sensitive ‘conditional’ ligands, thereby allowing the production of very large collections of pMHC complexes for T cell detection. Secondly, we have developed a ‘combinatorial coding’ strategy that allows parallel detection of dozens of different T cell populations within a single sample. The combined use of MHC ligand exchange and combinatorial coding allows the high-throughput dissection of disease- and therapy-induced CTL immunity. We have used this platform to monitor immune reactivity against a panel of 145 melanoma-associated epitopes in patients receiving Ipilimumab treatment. Results: Comparison of PBMC samples from 32 melanoma patients pre- and post-therapy indicated a significant increase in the number of detectable melanoma-associated T cell responses (p=0.004). Furthermore, kinetic data on T cell responses during therapy suggests that this broadening generally occurs within weeks after start of therapy. The magnitude of melanoma-specific T cell responses that was detectable prior to start of therapy was not significantly altered (p=0.8). Conclusions: These results establish the pattern of melanoma-specific T-cell reactivity induced by anti-CTLA4 treatment and form a benchmark for evaluation of other immunotherapeutic interventions, like anti-PD1 treatment, that are currently undergoing clinical evaluation. Furthermore, our data suggests that the clinical activity of Ipilimumab may be mostly due to epitope spreading, rather than through enhancement of pre-existing immune activity.